Meta's legal battle over fake ads reveal AI challenges in IP law

The field of intellectual property is accustomed to the nuanced application of existing legal frameworks to issues created by nascent and evolving technologies. Currently, artificial intelligence (AI) presents challenges to trademark, privacy, and copyright laws, such as whether AI-created images, videos, replicas, or voice simulations of actual people violate the Lanham Act as a false endorsement (15 U.S.C. §1125 (a)) or one's right of publicity under state or common law (Cal. Civ. Code §3344). Copyright law concerns arise with AI-generated digital replicas of artists and performers, imitations of their voice and likeness, and using and distributing copyrighted materials in the replicas or as part of a larger work. 17 U.S.C. §101, 106, et seq., and "Copyright and Artificial Intelligence, Part 1: Digital Replicas, A Report of the Register of Copyrights," U.S. Copyright Office (July 2024), available at https://www.copyright.gov/ai/Copyright-and-Artificial-Intelligence-Part-1-Digital-Replicas-Report.pdf.

Facebook and its parent Meta are implicated in the latest twist on AI content and digital replicas, raising a new issue as to whether it may be immune from liability for such information created and displayed on its platform under a 1996 law, Section 230 of the Communications Decency Act, which was intended to promote the development of the internet, yet provide immunity to an "interactive computer service", i.e., a social media platform, as to information posted on its service by another "information content provider." 47 U.S.C. §230(c)(1) "No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider."

In the action, Australian businessman and philanthropist Dr. Andrew Forrest argues that Meta's advertising tools have enabled scammers to create false advertisements using his likeness to promote fraudulent cryptocurrency scams. Forrest v. Meta Platforms, Inc., Case 5:22-CV-03699-PCP, (N.D. Cal. filed June 23, 2022). Multiple crypto scam advertisements using his likeness have been posted in Facebook over the years, and Forrest has tried many avenues to halt the creation and display of the scams, including hiring cybersecurity experts that scan Facebook and report any fraudulent advertisements to Facebook. Like Forrest, other finance executives such as Bill Ackman, Cathie Wood, and Steve Cohen have been the subject of financial scams posted on social media using their likeness with the goal of defrauding people out of money. Randall Smith, "Phony Billionaires on Facebook are Scamming Americans Out of Their Life Savings," Wall St. J. (March 15, 2024, 5:30AM), https://www.wsj.com/tech/fake-bill-ackman-cathie-wood-scam-a8df6ce7

Meta denies liability and argues that it is immune to the claims under Section 230 as an "interactive computer service" that merely provides a platform for the scammers to post the offensive material. As an interactive service, it is similar to a telephone provider that would not be liable for communications transmitted over it, or a bookstore selling offensive materials, however, Forrest argues that Meta shall be liable as it is acting as a digital advertising agency that contributes to the production of the ads. Particularly, it offers a Meta Ads program as a tool to run ads on platforms such as Facebook and Instagram, and bolsters the program with AI "Automation Tools" to optimize content creation. A question before the court is whether the use of these AI tools to create ads for others renders Meta an "information content provider" rather than simply a provider of an interactive computer service. If so, then Section 230 was not intended to immunize this type of conduct. Rigsby v. GoDaddy, Inc., 59 F.4th 998, 1007 (9th Cir. 2023).

Under the law, an "information content provider" is any entity "responsible, in whole or in part, for the creation or development of information provided through...any other interactive computer service." 47 U.S.C. § 203(c), (f) (emphasis added). If a provider makes a "material contribution to the creation or development of content," it may lose immunity. Kimzey v. Yelp! Inc., 836 F.3d 1263, 1269 (9th Cir. 2016). The court did not conclude whether Meta's advertising program, which involves the use of AI, creates or develops content in whole or in part, or contributes to the aspect of the content that is allegedly illegal. If the advertising program is deemed to create or develop any such content rather than simply perform as a publisher or speaker of entirely third-party content, Section 230 may not apply. Calise v. Meta Platforms, Inc., ---F.4th---, 2024 WL 2826231, at *9 (9th Cir. 2024). The matter is still pending as the court granted Meta's motion to dismiss certain causes of action in part, maintained Forrest's misappropriation and negligence claims, and left the factual issue regarding the mechanics of the advertising program and its creation or development of content for a later date.

If it can be shown that the content in the ad program was created and developed in whole or in part by Meta, or that the nature of its digital advertising program made a material contribution to the creation or development of the content posted, the next step will be for the court to address the misappropriation and negligence claims. This is important as it will address the intersection of AI digital replicas and content creation with intellectual property on a social media platform.

Additionally, if the court does not find immunity, future application of Section 230 with respect to AI generated content may be limited to scenarios where it can be shown that an interactive computer service such as a social media platform or content distributor is deemed to be an "information content provider" with respect to the offensive AI created content. Given the potential for the narrow application of whether a computer service is a content provider and the concurrent broad application of immunity from claims under Section 230, protection of one's rights will continue to be addressed via consistent monitoring, enforcement, licensing, and even litigation. Looking ahead, steps have been taken by the U.S. Copyright Office and policymakers to address the fast development of AI-driven content, as the absence of it could circumvent the overarching goals of our intellectual property laws to promote creation, reward innovation, protect consumers, and promote privacy and human rights. While financial scams are harmful to the public, even more destructive can be a lack of clear guidance as to the use and dissemination of content using AI that circumvents existing protections under existing intellectual property laws.